Preview: Phase II Study of AtuRNAi PF-04523655 for Wet AMD

PF-04523655, formerly known as RTP-801i, is the clinically most advanced RNAi Therapeutic candidate and has been in two phase II studies for diabetic macular edema (DME) and the exudative form of age-related macular degeneration (wet AMD). ‘655 is a 19bp blunt-end AtuRNAi trigger targeting the RTP801/REDD1 apoptotic stress response gene in the choroid. ‘655 was originally discovered by Silence Therapeutics and Quark Pharmaceuticals, licensed to Quark and eventually partnered by Pfizer which is largely in control of clinical development.

Data from the DME study were reported earlier this year (see related blog entry here). These showed that while ‘655 was well tolerated and efficacy was strongly suggestive of superiority versus laser photocoagulation, the old standard-of-care, Pfizer and Quark agreed to run a phase IIb study using higher dosages to take into account the emerging standard-of-care for DME, Lucentis, a monoclonal antibody against VEGF. Data from the phase II wet AMD study should be imminent.

Unlike the DME study, the ~150 patient wet AMD trial design for the wet AMD study, MONET, already took into account Lucentis as the new standard-of-care in that indication. As you can see from the treatment groups below, the goal is to either show superiority to Lucentis, or at least show a synergistic effect when used in combination with Lucentis (a positive outcome would be to show superiority of Arms 4 or 5 over Arm 1). Unlike many other wet AMD/DME-targeting agents in development, ‘655 should not act via the VEGF pathway and therefore has potential as an add-on to Lucentis.

Treatment arms in phase II wet AMD study:

Arm 1: 0.5 mg intravitreal injection of Lucentis given every 4 weeks from baseline to Week 12 (note: 0.5mg once a month is the recommended standard for Lucentis);

Arm 2: 0.5 mg of Lucentis given by intravitreal injection at baseline followed by 3 mg of PF-04523655 given every 2 weeks from Week 4 to Week 12;

Arm 3: 0.5 mg of Lucentis given by intravitreal injection at baseline followed by 1 mg of PF-04523655 given (weekly) from Week 4 to Week 12;

Arm 4: 0.5 mg of Lucentis given at baseline by intravitreal injection followed by 3 mg of PF-04523655 given every 4 weeks from Week 4 to Week 12;

Arm 5: 0.5 mg of Lucentis given by intravitreal injection at baseline followed by 1 mg of PF-04523655 (30 minutes later) given in combination every 4 weeks from baseline to Week 12.

Dosing had been completed and preliminary data from the study were known already in November 2010. Unfortunately, while ‘655 as single agent or in combination with Lucentis did show improvements in mean visual acuity over the 3 month dosing period, 6 and 9 letters respectively, it failed to show superiority to Lucentis at any of the doses at the important 4 month primary endpoint. No remarkable safety events were seen in MONET (all this can be gleaned from reading Quark Pharmaceutical’s latest prospectus). By comparison, large studies with Lucentis (0.5mg, monthly) have shown 6.6-9.8 letter average improvements from baseline at two years.

Outlook for ‘655 in Wet AMD

Most likely due to missing the primary 4-month endpoint, Quark Pharmaceuticals already indicated that following full data review Pfizer is unlikely to directly enter phase III trials, but would either decide to run a phase IIb study or abandon the program altogether. It is disappointing that ‘655 has apparently not shown synergism when used together with Lucentis as one might have expected from the presumed mechanism of action of ‘655, but possibly not from an immunostimulatory antiangiogenic VEGF-related artefact. A synergistic effect, however, may be masked by the initial Lucentis activity and only emerge over time. What is also often forgotten when it comes to the competitive space for the novel therapeutics in ocular diseases is long-term safety, not just absolute efficacy and dosing frequency, so studies with longer dosing than just the 3 months in MONET may be worth the risk. Maybe such safety and efficacy signals will emerge from the upcoming full data presentation.

Finally, instead of combining '655 with an antibody like Lucentis, combining it with the approved anti-VEGF aptamer Macugen (an oligonucleotide) might have practical advantages. Macugen has been struggling in the marketplace following the introduction of Lucentis due to perceived potency disadvantages. Curiously though, it is Pfizer that holds the commercial rights to Macugen, and combining it with '655 to reinvigorate the competitiveness of Macugen must have crossed their mind.

Quark RNAi Therapeutics Drug Candidate for Diabetic Macular Edema Effective, but Changing Competitive Landscape Weighs

A lot has changed in the treatment of diabetic macular edema (DME) since Quark sub-licensed its AtuRNAi compound to Pfizer in 2006. Most notably, the monoclonal antibodies against VEGF, Lucentis and the cheaper derivative Avastin, are replacing laser photocoagulation as the standard of care in this condition that adversely impacts the vision of up to 10% of those living with diabetes.

Quark last week reported 12-month data from the phase II study of PF-04523655 run by Pfizer (DEGAS) enrolling 184 patients with DME. The data suggest that the drug was on track of meeting its primary end-point which would have been superior improvements in mean visual acuity at 24 months compared to laser treatment: a 5.8 letter improvement for ‘655 at the highest dose (3mg) compared to 2.4 for laser (p= 0.08). Pfizer, the company that ran the trial, however, decided to terminate the study at this midway point as it deemed that the study objectives could no longer be met.

This is likely at least partly due to the new competition from the Roche/Novartis drug Lucentis that showed 24-month improvements of 12.5 letters versus 2.6 letters for sham control in a pivotal phase III study (RISE). Nevertheless, the efficacy of ‘655 looks much better if one considers that taking into account the 12 month completers only, the improvement was a very promising 9.1 letters for ‘655 versus 3.2 letters for laser. Moreover, the ‘655 data were reported to be dose responsive whereas the Lucentis data were not. Importantly, no serious adverse events were reported, although additional details remain to be reported at an upcoming conference.

Based on this dataset, it seems to me that ‘655 was efficacious, but unlikely to match the Lucentis data with the doses studied even when allowing for more patients to complete the study. Given the encouraging data, Quark decided to run a phase IIb study on its own dime, hoping that with higher doses the efficacy of Lucentis can be matched or even surpassed. This would confirm that the safety profile in DEGAS was satisfactory. In return, Quark would receive increased milestones and royalties should the drug eventually be approved.

Overall, the data are good news for RNAi Therapeutics in ocular indications in general. 3mg siRNA translate into almost a whopping 100mg per kilogram of human eye that were apparently well tolerated in this study (note, however, that the cost of goods for monoclonal antibodies would still be higher even at the current 0.3-0.5mg/injection). This paves the way for improved RNAi Therapeutics candidates with larger therapeutic indexes due to enhanced uptake and at the minimum the same frequency of administration as Lucentis (monthly). Some RNAi approaches, such as attempted by RXi Pharmaceuticals and Eyegate, may even dispense of the need for needle injections altogether. This would be a major differentiating factor.

Moreover, even with comparable single-drug efficacy, safety, and route of administration, ‘655 may be a viable alternative or complement to Lucentis as its mechanism of action is anticipated to be anti-apoptotic (stabilizing the blood-retina-barrier) instead of anti-angiogenic (preventing abnormal blood vessel formation). However, as indicated in my somewhat less optimistic blog entry last week, this is yet to be convincingly proven given the possibility that ‘655 may have at least partly worked through a non-RNAi mechanism. Demonstrating synergistic action in some animal models would go a long way in showing this.

For investors of Silence Therapeutics, I consider this better-than-expected data as they keep the drug alive and demonstrate the safety of the AtuRNAi design. How the new terms between Quark and Pfizer will affect Silence’s financial stake in the program and whether the present result will trigger any immediate payments, remains to be seen.

Update from March 18 Quark registration statement: In the parallel study of '655 in wet AMD, an interim analysis has shown improvements in mean visual acuity over 3 month period, but at 4 months, the primary endpoint, no numerical benefit over Lucentis was seen in any of the doses. The trial will be continued with full data expected in the second half of 2011.

Pfizer and Quark Progress RNAi Therapeutic for Diabetic Macular Edema into Phase II

Quark Pharmaceuticals today announced that development partner Pfizer has initiated dosing of the first patient in a phase II trial of the RNAi Therapeutic candidate PF-4523655/RTP801i-14) for diabetic macular edema (DME). The start of the phase II trials is a vote of confidence in RNAi Therapeutics for eye diseases by a major pharmaceutical company with a considerable eye care franchise and growing ambitions in RNAi Therapeutics.

The 160-person safety and efficacy trial follows phase I/II studies conducted by Quark that showed the drug to be safe and well tolerated in patients with wet age-related macular degeneration (wet AMD). Similar to DME which affects about 10% of type I and II diabetics, wet AMD is caused by the growth of leaky blood vessels in the back of the eye leading to blurred vision, and in some cases blindness. For this reason, it is not uncommon that a given compound is being tested for both indications.

The reason that Pfizer has chosen DME instead of wet AMD as the indication for this trial may be related to the fact that the wet AMD RNAi Therapeutics space has become crowded with Opko Health and Sirna Therapeutics/Merck having wet AMD RNAi Therapeutics in phase III and II studies, respectively. Opko Health also has a phase II study for DME using the same compound. It may also be related to the recent controversy caused by a Nature paper that showed that dsRNA-triggered immune responses alone may be anti-angiogenic in a mouse model for wet AMD and that this may have led to mis-interpretations of data from related pre-clinical RNAi Therapeutics studies.

However, PF-4523655 is unlike Sirna’s and Opko’s compounds in that it is intended to prevent abnormal blood vessel growth and leakage in a VEGF-independent manner. It is also a 19bp blunt-ended compound and therefore unlikely to trigger non-specific TLR3 responses (21bp seems to be the cut-off where you have to take TLR3 signaling into account). I trust that the former Coley Pharmaceutical’s folks with their deep understanding of TLR biology and that are now running the RNAi Therapeutics show at Pfizer will have closely looked at the issue.

Quark received an undisclosed amount for the initiation of the trials, Silence Therapeutics which had originally licensed the AtuRNAi technology to Quark collected $1.9M, and Alnylam from which a technology license was subsequently obtained on Pfizer’s insistence another undisclosed amount. Clearly, with today's allowance of Silence Therapeutics’ 10/633630 patent in the US which specifically claims AtuRNAi molecules of 18 or 19 base-pairs the IP battles are about to begin, and will be addressed in a follow-up blog entry. In any case, it is an interesting coincidence that the initiation of the phase II studies and the patent allowance occurred on the same day.

Independent of all these patent issues, I look forward to the results of this compound which has the makings of a safe and differentiated RNAi Therapeutic for DME.